Lexical Acquisition through Acquisition order

It is often noted in the literature that typical instances of categories denoted by lexical items are acquired earlier than atypical instances. Yet, a line of empirical studies that were left relatively unnoticed provide evidence to an additional generalization. These studies show that early acquired members are often interpreted as typical of their category, thereby forming a basis for generalizations concerning the dimensions identifying category members. Based on this evidence, I propose a new intra-domain bootstraping mechanism, which I call the Learning Bias. I propose that the bootstrapping of the meaning of simple and complex category type expressions like bird, tall and tall bird, is based on the order in which their early acquired members are classified. I review existing evidence supporting the hypothesis that the Learning Bias is at work in children of different ages, of both typical and atypical populations, and it remains operative in adults. The Learning Bias affects both linguistic and nonlinguistic categories, therefore being a domain-general learning mechanism. Finally, I propose that the reviewed data indirectly supports the lexical bootstrapping hypothesis. Along the paper, I draw attention to open questions and details, and provide concrete proposals for future research on the topics discussed.